Alzheimer’s disease (AD) is the most common cause of dementia in people over the age of 65y. It is a progressive neurodegenerative disorder that can manifest as deficits in memory, executive function, visuospatial cognition, language function, and personality changes(Albert et al. 2011; McKhann et al. 2011). While a clinical diagnosis of probable AD dementia can be made if other causes of dementia are ruled out, a definitive diagnosis of AD neuropathological changes requires a histological examination of brain tissue. Key neuropathological features of AD include neurofibrillary tangles (NFTs), primarily composed of abnormally phosphorylated tau protein, extracellular deposition of amyloid-ß peptides(Aß) in senile plaques, and dystrophic neurites(Hyman et al. 2012; Montine et al. 2012).
The amyloid hypothesis of AD postulates that an imbalance between Aß production and clearance, resulting in the accumulation of Aß is a driver of AD pathogenesis(Hardy and Selkoe 2002). While rare genetic early-onset forms of AD are associated with the over-production of abnormal proteins, including Aß and tau(Scheuner et al. 1996; Naj et al. 2017), the more common late-onset Alzheimer’s disease is associated with reduced clearance of Aß from the brain(Mawuenyega et al. 2010; Tarasoff-Conway et al. 2015; Zuroff et al. 2017). However, therapies utilizing monoclonal antibodies targeting Aß to increase efflux from the brain have been largely unsuccessful(Salloway et al. 2014; Doody et al. 2014), indicating that AD pathogenesis is not so simple.
AD is a complex disease, and its etiology is likely multifactorial. There is a growing body of work which links neuroinflammation, oxidative damage, and dysfunctional glucose and lipid metabolism to AD(Kapogiannis and Mattson 2011; Croteau et al. 2018; Butterfield and Halliwell 2019b, a; Behl et al. 2020; Leng and Edison 2020; Ionescu-Tucker and Cotman 2021; Meng et al. 2022). Accordingly, diabetes mellitus and hypercholesterolemia are major risk factors for developing AD(Shepardson et al. 2011a; 2013; Arnold et al. 2018; Meng et al. 2022). Patients with AD have brain insulin resistance(Arnold et al. 2018), and treatment with intranasal insulin improves cognitive function(Reger et al. 2008). Meanwhile, cholesterol has been shown to accumulate in senile plaques(Mori et al. 2001) and increases production of Aß(Shepardson et al. 2011a), but evidence on the effects of statins on cognitive function is mixed(Kurata et al. 2011; Sano et al. 2011; Shepardson et al. 2011b; Tong et al. 2012). Nevertheless, Apolipoprotein E4 remains the most potent genetic risk factor for the development of AD(Naj et al. 2017). Interestingly, of the identified transporters that efflux Aß out of the brain, many, like LRP1, LRP2, ABCA1, ABCB1 (also known as P-glycoprotein-1), and ABCG4(Shibata et al. 2000; Cirrito et al. 2005; Bell et al. 2007; Do et al. 2012; Dodacki et al. 2017), also have roles in cholesterol metabolism.
ABCG4 is a member of the ATP-binding cassette transporter family that regulates cholesterol homeostasis; ABCG4 is predominantly expressed in the CNS(Yoshikawa et al. 2002; Cserepes et al. 2004; Bojanic et al. 2010). It has also been suggested to have a function in glucose-stimulated insulin secretion (GSIS)(Hou et al. 2016). In the brain, ABCG4 is expressed in neurons, astrocytes, microglia, and capillary endothelial cells at the blood-brain barrier(BBB) (Tarr and Edwards 2008; Uehara et al. 2008; Bojanic et al. 2010; Dodacki et al. 2017). Abcg4 may play a role in cholesterol efflux from the brain(Wang et al. 2004, 2008; Vaughan and Oram 2006), and in vitro studies suggest it may also play a role in the export of Aß from the brain at the BBB (Do et al. 2012; Dodacki et al. 2017). Furthermore, in in vitro studies, ABCG4 was found to inhibit γ-secretase activity, thus reducing Aß production(Sano et al. 2016). Abcg4−/− mice were reported to have a deficit in contextual memory(Bojanic et al. 2010) though no other confirmatory reports have been published. It has therefore been posited that Abcg4 may play a protective role against the development of AD.
We sought to test the hypothesis that Abcg4 may be involved in AD pathogenesis using Abcg4 knockout (KO) mice. We chose the APPSwe,Ind (also referred to as J9) mouse model of AD because it has been reported to have a slower onset AD than other models, and loss of Abcg4 on this background would be expected to accelerate onset, should the hypothesis be supported. The J9 model is a transgenic mouse that expresses human amyloid precursor protein (APP) with Swedish (K670N/M671L) and Indiana (V717F) mutations, which increases Aß formation and favors Aß42, the form more likely to be found in senile plaques(Hsia et al. 1999; Mucke et al. 2000). We crossed J9 mice with Abcg4 KO mice and assessed metabolic and behavioral effects.